cAMP-mediated autophagy inhibits DNA damage-induced death of leukemia cells independent of p53-article summary

(1)

cAMP-mediated autophagy inhibits DNA damage-induced death of

leukemia cells independent of p53

Article summary

Eva Duthil

Project thesis ved Institutt for medisinske basalfag UNIVERSITETET I OSLO

Februar 2019

(2)

ii

ABSTRACT

Autophagy is important in regulating the balance between cell death and survival, with the tumor suppressor p53 as one of the key components in this interplay. We have previously utilized an in vitro model of the most common form of childhood cancer, B cell precursor acute lymphoblastic leukemia (BCP-ALL), to show that activation of the cAMP signaling pathway inhibits p53-

mediated apoptosis in response to DNA damage in both cell lines and primary leukemic cells. The present study reveals that cAMP-mediated survival of BCP-ALL cells exposed to DNA damaging agents, involves a critical and p53-independent enhancement of autophagy. Although autophagy generally is regarded as a survival mechanism, DNA damage-induced apoptosis has been linked both to enhanced and reduced levels of autophagy. Here we show that exposure of BCP-ALL cells to irradiation or cytotoxic drugs triggers autophagy and cell death in a p53-dependent manner.

Stimulation of the cAMP signaling pathway further augments autophagy and inhibits the DNA damage-induced cell death concomitant with reduced nuclear levels of p53. Knocking-down the levels of p53 reduced the irradiation-induced autophagy and cell death, but had no effect on the cAMP-mediated autophagy. Moreover, prevention of autophagy by bafilomycin A1 or by the ULK- inhibitor MRT68921, diminished the protecting effect of cAMP signaling on DNA damage-induced cell death. Having previously proposed a role of the cAMP signaling pathway in development and treatment of BCP-ALLs, we here suggest that inhibitors of autophagy may improve current DNA damage-based therapy of BCP-ALL - independent of p53

(3)

iii

1. INTRODUCTION

1.1 B cell precursor acute lymphoblastic leukemia.

B cell precursor acute lymphoblastic leukemia (BCP-ALL, hereafter named ALL) is the most common form of cancer in children. It accounts for 80% of pediatric leukemias, and it is a type of cancer originating from B cell precursors in the bone marrow. The etiology is unknown, but there are disposing factors like Downs syndrome, immunodeficiencies and high levels of radiation in utero.

Incidence of ALL peaks at age 2-5 years. Presentation of ALL in children can manifest in several ways and not necessarily with an acute start. Common symptoms include signs of reduced general condition, increasing fatigue, anorexia, reduced activity, muscle and bone pains and an ill-looking appearance. Many symptoms can be ascribed myelosuppression, that result in anaemia, neutropenia, and thrombocytopenia, clinically manifested as pallor and lethargy. Infections, fever, lymphade- nopathy, bruising and nosebleeds are common. Like most cancer forms, ALL has the potential to disseminate to other organs. Infiltration of reticulo-endothelial organs leads to hepatosplenomegaly with abdominal pains, whereas dyspnea occurs upon lung-infiltration. Central nervous system (CNS)-leukemia, although rarely present at diagnosis, most often occurs at relapse and can give headache, neck-stiffness, vomiting and other neurological symptoms (1).

Diagnosis of ALL is made by direct bone-marrow analysis, upon suspicion from the clinical presentation and an indicative blood status (anaemia, neutropenia and thrombocytopenia). Finding >25%

blasts in a bone marrow sample indicates ALL. Immunophenotyping and cytogenetic analysis are made to further sub-classify the type of ALL. Patients with ALL are risk-stratified according to the WHO classification system for acute leukemias, and the treatment administered is adjusted accordingly. The risk groups are divided into low-, average-, high-, very high-, and special groups based on the number of white blood cells detected at diagnosis, response to therapy, infiltration of CNS or testes and cytogenetic profile.(2)

Treatment is divided into three phases; induction, consolidation and sustaining. Prognosis is influ- enced by the age of the patient, tumour load, cytogenetic abnormalities, response to initial chemotherapy and minimal residual disease assessment (MRD)(3). The latter is detected by polymerase chain reaction (PCR) after induction therapy. Correct classification is important to overcome therapy-resistance and risk of relapse, especially for the poor prognostic subgroups of ALL.

(5)

2

Thanks to improved combinational chemotherapeutic treatment over the last 20 years, the prognosis is generally good for children with ALL. The 5-year survival is now close to 90 %, although the prognosis varies with different cytogenetical subgroups. For instance is hyperdiploidity (> 50 chromosomes) a good prognostic indicator, whereas hypodiploidity (<44 chromosomes) is associated with poorer outcomes. Structural abnormalities such as translocations, deletions, insertions and in- versions can cause instability, influencing the level of malignancy. The t(12;21) translocation creat- ing the fusion-gene ETV6-RUNX1 is favourable, whereas the Philadelphia chromosome t(9;22) BCR/ABL1 rearrangement, the t(4;11)KMT2A (MLL) rearrangement and iAMP21 (intrachromo- somal amplification of chromosome 21) are considered as unfavourable features (4).

The treatment of ALL is extensive, with a duration of two and a half years. In Norway all children diagnosed with ALL are treated according to common nordic guidelines, the NOPHO-protocol (Nordisk forening for pediatrisk hematologi og onkologi (5)

Treatment aims to obtain rapid and continuous remission while minimizing the side effects of the treatment. Late complications related to exposure to chemotherapy and radiation are common, often presenting between 10-40 years after treatment (3). The CNS is especially vulnerable leading to conditions such as reduced cognitive function and psychiatric diseases, motor disturbances, hearing loss and damage to the eyes. Other common side-effects include heart-, lung- and metabolic diseases, such as lack of growth hormones, overweight, reduced fertility, osteoporosis, odontological conditions, and secondary malignancies. Following up and preventing the late effects in patients receiv- ing treatment is important (3).

1.2 DNA-damage responses

Conventional treatment of ALL is multimodal chemotherapy. The main antineoplastic drugs used are vincristine, doxorubicin, 6-mercaptopurine and methotrexate. The vinca-alkaloid vincristine is an anti-mitotic agent that inhibits metaphase by binding to tubulin protein leading the cell into apoptosis. Doxorubicin is an anthracyklin inhibitor that intercalates DNA and inhibits topoisomer- ase-II leading to DNA damage. 6-Mercaptopurine is a purine antagonist and inhibits DNA and RNA synthesis. Methotrexate is a folate antimetabolite and inhibits that synthesis, repair of DNA and cellular replication.(6) DNA damaging irradiation is used only for treating CNS infiltrations in pediatric ALL patients (5).

Cells have specialized repair pathways to help maintaining genome integrity, and DNA damaging chemotherapy and irradiation must overcome these barriers to kill the cancer cells. DNA lesions

(6)

3

trigger multiple signaling-cascades, collectively termed DNA damage response (DDR), that regu- late activation of cell cycle checkpoints and trigger arrest, repair and resumption or even result in cellular apoptosis if the damage persists (7). When DNA is damaged, for instance by ionizing radiation causing double strand breaks (DSB), the serine threonine kinases ATM (ataxia telangiectasia mutated)-protein or ATR (ataxia telangiectasia and rad3-related protein) are recruited and activated.

They will then activate checkpoint kinases 1 or 2 (Chk1/2), that in turn will phosphorylate a number of substrates. Among these targets is the p53 protein, which will then change conformation from an inactive HDM2-bound state and accumulate within the cell. Usually, p53 is bound to HDM2, and targeted for ubiquitination and proteosomal degradation; therefore the p53 levels remain low. The p53 protein is an important regulator in the DDR-response. It is a tumor suppressor also called

“guardian of the genome”, because of its decisive role in maintaining genome integrity (8). P53 will halt the cell cycle until DNA breaks are repaired, it will induce the DNA-repair-machinery, and it will induce pro-apoptotic responses if the DNA damage is not repaired (9). All of these effects are mediated by nuclear p53 acting as a transcription factor.

Figure 1: Schematic presentation of the DNA-damage response. (10)

(7)

4

The DNA-damage response (DDR) is aberrantly regulated in cancer cells, often due to mutations in the genes encoding the DDR components (11). The p53 gene itself is for instance mutated in more than 50 % of cancers (12), resulting in cells surviving with chromosomal aberration.(13) Subtypes of paediatric ALLs present with wild-type p53 (14). P53 protein levels are kept low due to overex- pression of the E3 ligase HDM2 resulting in proteasomal degradation of the p53 protein. Treatment resistance in ALL is associated with a phenotype of non-functional p53 and P53 mutations are fre- quent finding in relapsed cases (15, 16).

1.3 Autophagy - guardian of the proteome

Besides p53, autophagy has profound effects on cancer cell survival. It is a physiological process with the purpose of maintaining cellular homeostasis by removal of damaged proteins and organelles in cells exposed to stressful conditions. (17) Initiation of a membraneous structure, the phagophore, grows and becomes a double-membraned autophagosome. (17) Cell organelles and proteins targeted for degradation are encapsulated into the autophagosome, which then fuses with lysosomes. The content is finally degraded, and recycled or catabolized to produce ATP for the cell.

This is beneficial for cell-survival for instance during nutrient deprived periods. Much as in the same way as p53 maintains genome integrity, autophagy maintains proteome integrity. (18)

Figure 2: Schematic illustration of autophagy (19)

The role of autophagy in cancer development is complex and has been controversial. The consensus seems now to be that whereas autophagy prevents the initiation of cancer, it paradoxically will pro-

(8)

5

mote the progression of cancer at later stages. Therefore, targeting autophagy in the context of preventing and treating cancer is considered as a “double-edged sword” (19, 20)

Although autophagy generally is regarded as a cellular survival mechanism, extensive autophagy may also trigger so called “autophagic” cell death, for instance in rapid growing tumors susceptible to lack of nutrients and oxygen. (21) In the context of BCP-ALL, subtypes having the BCR-ABL1 fusion protein display low basal autophagy but are highly dependent of it for their survival. Therapy resistance related to for instance glucocorticoid treatment of BCP-ALL has also been linked to activation of autophagy (22).

1.3.1 The role of p53 in autophagy

The tumor suppressor p53 has also a pivotal role in controlling autophagy, by regulating the transcription of multiple genes involved in the autophagy process. (23) For instance will the p53 target gene DRAM (damage-regulated autophagy modulator) that is associated with p53-mediated apoptosis, also induce autophagy (24). However, as a key protein in DDR, the role of p53 in DDR-

mediated autophagy has been debated. It seems that p53 controls autophagy in a complex manner, which depends on its subcellular localization and involves transcriptional and non-transcriptional mechanisms. The consensus seems to be that while nuclear p53 stimulates the autophagic program (thereby sustaining the attempts of cells to cope with stresses such as DNA damage), cytoplasmic p53 inhibits autophagy and hence facilitates cell death. (25) TP53-induced autophagy has also been demonstrated in malignant lymphocytes, including in BCR/ABL-positive BCP-ALL. (26) Upon treatment with DNA damaging anti-cancer agents, wt p53 will either induce cell cycle arrest, DNA repair, apoptosis or autophagy. It appears that the outcome of the treatment will depend on the balance between these processes, and that the autophagy mode is generally associated with worse clinical outcome. (26)

1.4 The cAMP-signaling pathway

Cyclic AMP is formed by G-protein-coupled receptor (GPCR)-mediated activation of adenylyl cyclases. (27) GPCRs are a large family of receptors that respond to extracellular stimuli ranging from small molecules like catecholamines (epinephrine/norepinephrine) to large hormones. (28) cAMP activates a family of cAMP-dependent kinases (PKAs), leading to phosphorylation of proteins regulating important cell features like metabolism, proliferation and apoptosis. In lymphocytes, substances like catecholamines, serotonin, histamines and prostaglandin E2 (PGE2) activate the cAMP/PKA pathway. (29)

(9)

6

Increasing evidence supports the notion that development of solid cancers, including B cell lym- phomas, involves the interplay between the tumour cells and the microenvironment. (30, 31) How- ever, less is known about this interaction in the context of leukemias. It has been demonstrated that cancer cell-stimulated mesenchymal stem cells are able to form a carcinoma stem cell niche via autocrine production of the cAMP-elevating compound PGE2, (32) and it has been shown that mesenchymal stem cells in the bone marrow secrete PGE2 in a cyclooxygenase-dependent manner. (32, 33) In addition to supporting the viability and function of normal hematopoietic cells, the bone marrow microenvironment is also believed to provide vital support to leukemic blasts. (34-36) BCP- ALL cells typically develop in close contact with stromal cells in the bone marrow. The bone marrow microenvironment also forms niches that can confer protection against damage induced by therapeutic agents, thereby contributing to development of treatment resistance. (37)

Our lab previously proposed that PGE2 secreted from bone marrow stromal cells could contribute to development of BCP-ALL and to treatment resistance by activating the cAMP-pathway via EP2 receptors on the surface of the leukemic cells. (38) Using a co-culture system between bone marrow stromal cells and primary ALL cells from patients, it was demonstrated that the stromal cells in such co-cultures indeed secrete PGE2. (39) Furthermore is was shown that both p53 levels and apoptosis induced by irradiation in the ALL cells were reduced by approximately 50-60 % when co- cultured in the presence of the bone marrow-derived stromal cells. (39) Production of PGE2 is cata- lysed by cyclooxygenase. As presented in the model in the figure below, we used the cox-inhibitor indomethacin to show that PGE2 produced by the stromal cells indeed contribute to the counteract- ing effect on DNA damage-induced p53 and apoptosis in BCP-ALL. (39)

(10)

7

Figure 4: Schematic model for MSC derived PGE2 mediated cytoprotection of BCP-ALL (39)

2. MOTIVATION AND AIMS OF THE PRESENT PROJECT

2.1 Motivation for the present work

Suppression of normal p53 function is assumed as a prerequisite for malignant cell

development.(40) However, in contrast to most cancers, mutations in the p53 gene are relatively rare in hematological malignancies. It is therefore reasonable to assume that cancers such as ALL depend on alternative mechanisms to mitigate p53 functions. DNA damage responses are the basis for common anti-cancer treatments like gamma-irradiation and chemotherapy, but many patients relapse after initial treatments. In order to increase the efficiency of a given treatment and to limit side effects on normal cells, there is a constant search for modulators of DNA-damage responses as targets in cancer therapy. We believe that our BCP-ALL model provides a unique opportunity to reveal the role of both autophagy as such, and p53 in particular, in DDRs.

2.2. Aims

The aim of the current project was to explore the role of autophagy in cAMP-mediated protection of DNA damage-induced cell death.

3. RESULTS

The results of the present project are presented in the current paper.

4. DISCUSSION

4.1 Methods for analysis of autophagy

A common way to monitor the dynamic process of autophagy is by measuring the formation of autophagosomes in a cell, this dynamic process is known as autophagic flux. As mentioned previously, autophagosomes are double-membraned vesicles with cellular content targeted for degradation.

During autophagy the protein named microtubule-associated light chain 3 (LC3) is recruited to the autophagosomes. The cytosolic form LC3-I is conjugated to phosphatidylethanolamine (PE) form- ing membrane-bound LC3-II. Intra-autophagosomal LC3-II is degraded at the same time as the content is broken down by acidic hydrolases after lysososomal fusion. Changes in LC3-II levels corre- late with the degree of autophagosome formation and reflect autophagic flux.

(11)

8

Figure 5: Schematic overview of cytosolic and membrane-bound LC3 during autophagy (41) Considering that LC3-II is degraded in lysosomes, it can be difficult to interpret if an increase rep- resents induction of autophagy or simply a degradation block. Increasing autophagic activity would result in higher LC3-II levels as would compromised degradation, for instance because of delayed lysosomal fusion. To exclude the possibility of degradation-block as the cause of LC3-II increase, lysosomal protease inhibitors are used. BafilomycinA1 (BafA1) is a compound that targets vacu- loar-type H+ATPase enzyme, in the context of autophagy BafA1 prevents acidification of lysosome so that the proteases remain inactive. In addition, it inhibits autophagosome-lysosome fusion.

Figure 6: Schematic illustration of target actions of bafilomycin (42)

Described below is a systematic overview of the different methods performed for analysis of autophagy.

(12)

9

4.1.1 LC3-II/I-ratio normalized by loading controls on western blot

The experiments measuring autophagic flux in our study where performed both with and without BafA1, using cultured ALL cell line REH. Intracellular cAMP-signaling was triggered by adding forskolin, a direct activator of cAMP(43), to the cell-cultures, prior to inducing DNA-damage by X- rays (IR). Autophagic flux was finally analyzed by western blot after 6 and 24 hours, portraying separate bands at different molecular weights for LC3-II and LC3-I. Figure 6 shows how the cAMP-activator forskolin enhances the conversion of LC3I to II in REH cells both in the absence and presence DNA damaging irradiation. cAMP-signaling activation and IR-treatment independently increased the autophagic flux in REH, and had a noticeable synergistic effect on autophagic flux.

Hence, enhancing the levels of cAMP promotes autophagy.

Figure 7: Autophagic flux by western blotting showing differences in intensity of LC3-II and LC3-I bands 24 hours post irradiation, normalized to calnexin loading controls (left panel). Quantification of western blot as LC3-II/LC3-I ratios for the different intervention groups (right panel). Autophag- ic enhancement after treatment with BafA1, added for the last 4 hours, suggests a proper increase in autophagy.(44)

Increasing intracellular levels of cAMP stimulated autophagy, but we also wanted to demonstrate it occurred by downstream PKA-enzyme activation. The similar experiment using PKA-activator (8- CPT) had the same effect as using forskolin (AC-activator) on autophagy. In addition, inhibition of PKA-activation, attenuated the autophagy enhancement of forskolin and IR-treatment. Advantages of LC3-II/LC3-I ratio normalized by loading controls on western blot, is that it discriminates well between the lipidated form, LC3-II and LC3-I. LC3-II is also specifically associated to autophagosome, well reflecting the number quantity autophagosomes in a cell. A disadvantage is that autophagic flux is a dynamic process whereas this methods only portrays a still image of this process.

There is no standardized answer to the rate of occurrences for different cell-lines and interventions.

Determining the accurate time point, can be challenging when lacking knowledge of how this process occurs in time, and requires kinetic-specific experiments to figure out how the LC3-II an LC3-I develop over time for a particular experimental setup.(45)

(13)

10

4.1.2 LC3B puncta analysis

Other methods were also used to measure autophagic flux in REH-cells, we analyzed the formation

of autophagosomes by visualizing LC3-puncta with confocal microscopy using immune- histochemistry. Cell-cultures with exposures as described above were fixed after 24 hours and care-

fully centrifuged so that the cells project onto glass slides, by a method called cytospin. The cells were then permeabilized and stained with fluorescent labeled LC3-specific antibodies. Viewed under a fluorescence-detecting confocal microscope, autophagosomes appear as dots or “puncta”, and are quantified by counting. Cells that had been exposed to irradiation or forskolin expressed a higher amount of puncta. As shown in figure 7, we observed that forskolin enhanced the number of LC3 puncta both alone and in the presence of irradiation.

Figure 8: Confocal images of REH stained for LC3. (44)

This method is adequate for visualizing selected intracellular compartments. Confocal microscopy enables a more precise three dimentional approach as one can scroll through transverse sections throughout the entire cell. LC3 immunostaining of fixed cells does not exceptionally stain mem-

(14)

11

brane bound lipidated LC3-II. Background fluorescence from LC3-I can create the impression of falsely increased autophagosome numbers. To minimize this problem a permeabilisation process is performed, in order to wash away unbound LC3, but also increases the risk of disrupting the autophagosome membranes.

4.1.3 CYTO-ID staining

To validate that the puncta were in fact specific for autophagosome membrane, additional CYTO- ID staining was performed. This compound stain autophagic vesicles selectively. We could thereby confirm co-localisation of CYTO-ID and LC3-puncta by confocal.

Figure 9: IR and forskolin-treated REH cells stained with LC3B and CYTO-ID(46)

Advantages of CYTO-ID- easy to quantify intensity by flow cytometry. Can also be used in live cells and does not require any permeabilisaton process like the previous method. We further pro- ceeded with a more accurate quantification of autophagy by CYTO-ID on flow cytometry.

This technique requires lysosmal inhibition, because autophagic vesicles are degraded in autoph- agolysosomes. They are also partially specific and present some unspecific background.(47) 4.1.4 Degradation of long lived proteins

Constituent intracellular macromolecules like long lived proteins are in large proteolysed by autophagy. Degradation of long lived proteins is used as an assay of autophagic degradation. Short lived proteins have a high turnover and are proteolysed more rapidly than the LLP who later under- go autophagic degradation.(48) LLP can be labelled by radioactive amino acids, later release of radioactivity from the labelled protein is measured . Autophagic degradation by turnover of radioactive labelled LLP is interpreted as autophagic flux. We found that irradiated and forskolin as well as forskolin-only treated cells presented with a higher long lived protein degradation rate. This method is a direct method to measure autophagic flux. One disadvantage is that LLPD assays are best per-

(15)

12

formed using starvation medium. Starvation medium is used as an autophagy inducer in other experimental setups. Proteasomal degradation of LLPD can still occur in this assay, but is minimized by using proteasome inhibitors or performing the experiment with the presence and absence of autophagy inhibitors(47)

Figure 10: quantification of degradation of long-lived protein assay for IR and forskolin treated REH(46)

4.1.5 siRNA of ULK1

ULK1 is a protein embedded in the ULK1-complex, needed in the early steps of autophagy. It is encoded by the ULK1-gene. We inhibited ULK1 expression by targeting its mRNA, this is a way of knocking down this protein. This was is done by silencer expressing plasmid transfection. REH- cells were transfected with small interfering oligonucleotides (siRNA) targeting ULK1 and non- targeting siRNA as control. Forskolin induced CYTO-ID staining was prevented by knock-down of ULK1 RNA. This technique made it possible for us to gain insight in upstream autophagy modulation. The technical challenges rely on the transient process, and not fully inhibiting ULK1.

4.1.6 Real time PCR transcription analysis (MAP1LC3B and SQSTM1)

MAP1LC3B is a gene encoding the protein LC3. To analyse expression of this gene, we used the principle of reverse transcription of messenger RNA into complementary DNA (cDNA). An oli- goprimer binds to mRNA and the enzyme reverse transcriptase bind to this primer and synthesises complementary strand, the cDNA. The latter being amplified by real time PCR. Specific binding dye is introduced and fluorescence signal will increase proportionally to the amount of DNA repli- cated. The aim is to measure the amount of specific RNA transcribed from a gene of interest. 24 hours after exposing REH cell to irradiation and forskolin, RNA was isolated. cDNA was then cre- ated by reverse transcription and amplified. In this context, the mRNA levels of MAP1LC3B were significantly increased in REH cells co-treated with forskolin and irradiation. REH treated with forskolin exempt irradiation or the opposite had only subtle increased RNA levels. For this technique,

(16)

13

RNA has to be removed from the cell, through a meticulous purification process. When using housekeeping genes as reference genes for control, one assumes they are not being up or down regulated, and relies on same expression pattern in all intervention groups.(49)

4.2. Discussion of the results

Firstly we confirmed that IR augmented levels of p53. Next, siRNA knock-down of p53 reduced the levels of IR-induced autophagy and cell death. Surprisingly, knocking down p53 did not have implications for the forskolin-induced autophagy after IR. We concluded that the interplay between cAMP and autophagy is independent of p53.

Based on the findings in our group, we suggest that cAMP-signaling inhibits DNA damage-induced cell death by at least two different mechanisms. We have previously shown that activation of the cAMP/PKA pathway reduces the level of DNA damage-induced p53, resulting in reduced apoptosis. In the present study, we have shown that cAMP-signaling promotes the survival of DNA damaged ALL cells by enhancing the level of autophagy in a p53-independent manner. We propose that autophagy has to exceed a certain threshold to enable cell survival, illustrated in figure 8. Irradiation promotes a low level of autophagy in an attempt to survive the DNA damaging injury. However, the level is too low for the cells to survive. According to our model, cAMP-signaling will enhance the level of autophagy above the threshold required to survive. Knowing that the tumor microenvironment secretes PGE2 which generates cAMP in BCP-ALL, this can explain how autophagy could be promoted above this threshold in vivo, making the cells more resilient to DNA-damaging therapy.

Noninvasive modalities and methods to accurately measure autophagy and interpret its significance in pathological processes required for in vivo studies are a current work in progress. A clinical aim is to optimize current cytotoxic treatment by making it more targeted toward the pathological process and mitigate side-effects. We believe that targeting the cAMP-signaling pathway in connection with DNA-damaging therapy of ALL, could be a way of enhancing the effects of the treatment.

(17)

14

Figure 9: “Proposed model for cAMP-mediated survival of DNA damage-induced BCP-ALL cells.”(44)

5. CONCLUSIONS

Increasing awareness of the implications of autophagy in relation to cancer development and its interaction with key-pathways of DNA-damage responses, lead us to gain further insight into autophagy-related mechanisms occurring in BCP-ALL. We looked at the interactions between DNA- damage responses, apoptosis and autophagy in BCP-ALLs, in the context of cAMP-mediated signalling.

BCP-ALL often has wild-type p53, but suppression of p53 accumulation still occurs through cAMP-mediated signalling, facilitating survival through compromising the apoptosis response after DNA-damage. In the present project, we have elucidated the mechanisms whereby cAMP-signaling protects ALL cells from DNA damage-induced cell death. In particular, we have explored the role of autophagy in this process. Induction of cAMP-signalling stimulated autophagy providing additional prosurvival and protective effects. Still, blocking autophagy had minimal impacts on IR- induced killing, and we believe the levels of autophagy in the BCP-ALL are too low to enable protection. On the other hand, when the cells were deprived of the possibility to induce autophagy, the

(18)

15

cAMP-mediated protective effects were diminished. Based on our findings, we suggest that autophagy has to exceed a certain threshold to enable protection. Knowing that the tumor microenvironment secretes PGE2 whom generates cAMP in BCP-ALL, this can explain how autophagy could be promoted above this threshold in vivo, making the cells more resilient to DNA-damaging therapy.

Targeting the cAMP-signaling pathway in conjunction to DNA-damaging therapy could be a way of compromising the BCP-ALLs survival.

(19)

16

6. REFERENCES

1. Illustrated Textbook of Paediatrics 5th Edition. In: Lissauer Tom CW, editor.; 2017:

Elsevier; 2017. p. 600.

2. Terzah M Horton M, PhD, C Philip Steuber, MD, Jon C Aster, MD. Overview of the clinical presentation and diagnosis of acute lymphoblastic leukemia/lymphoma in children

uptodate.com2019 [Available from: https://www.uptodate.com/contents/overview-of-the-clinical- presentation-and-diagnosis-of-acute-lymphoblastic-leukemia-lymphoma-in-children#H2898462587 3. Leukemi hos barn: Norsk Elektronisk Legehåndbok; [updated 11.07.2018. Available from:

https://legehandboka.no/handboken/kliniske-kapitler/pediatri/tilstander-og- sykdommer/kreftsykdommer/leukemi-hos-barn/#reference-3.

4. Terzah M Horton M, PhD, C Philip Steuber M. Risk group stratification and prognosis for acute lymphoblastic leukemia in children and adolescents UpToDate2018 [Available from:

https://www.uptodate.com/contents/risk-group-stratification-and-prognosis-for-acute- lymphoblastic-leukemia-in-children-and-

adolescents?sectionName=Structural%20abnormalities&topicRef=4686&anchor=H8&source=see_l ink#H8.

5. Nasjonalt handlingsprogram med retningslinjer for diagnostikk,

behandling og oppfølging av kreft hos barn. In: Helsedirektoratet, editor. 3 ed: Helsedirektoratet;

2017.

6. General drug information uptodate.com2019 [Available from:

https://www.uptodate.com/contents/table-of-contents/drug-information/general-drug-information.

7. O’Connor Mark J. Targeting the DNA Damage Response in Cancer. Molecular Cell.

2015;60(4):547-60.

8. Lane DP. Cancer. p53, guardian of the genome. Nature. 1992;358:15-6.

9. Ljungman M. Activation of DNA damage signaling. Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis. 2005;577(1):203-16.

10. Sulli G, Di Micco R, di Fagagna FdA. Crosstalk between chromatin state and DNA damage response in cellular senescence and cancer. Nature Reviews Cancer. 2012;12:709.

11. Broustas CG, Lieberman HB. DNA damage response genes and the development of cancer metastasis. Radiation research. 2014;181(2):111-30.

12. Vogelstein B, Lane D, Levine AJ. Surfing the p53 network. Nature. 2000;408:307.

13. Schär P. Spontaneous DNA Damage, Genome Instability, and Cancer—When DNA Replication Escapes Control. Cell. 2001;104(3):329-32.

14. Pinto EM, Ribeiro RC, Figueiredo BC, Zambetti GP. TP53-Associated Pediatric Malignancies. Genes & cancer. 2011;2(4):485-90.

15. Hanahan D, Weinberg RA. The Hallmarks of Cancer. Cell. 2000;100(1):57-70.

16. Levine AJ. p53, the Cellular Gatekeeper for Growth and Division. Cell. 1997;88(3):323-31.

17. Klionsky DJ, Emr SD. Autophagy as a Regulated Pathway of Cellular Degradation. Science (New York, NY). 2000;290(5497):1717.

18. Gumeni S, Evangelakou Z, Gorgoulis VG, Trougakos IP. Proteome Stability as a Key Factor of Genome Integrity. International journal of molecular sciences. 2017;18(10):2036.

19. Ryan KM. p53 and autophagy in cancer: Guardian of the genome meets guardian of the proteome. European Journal of Cancer. 2011;47(1):44-50.

20. Wilkinson S, Ryan KM. Autophagy: an adaptable modifier of tumourigenesis. Current Opinion in Genetics & Development. 2010;20(1):57-64.

21. Tsujimoto Y, Shimizu S. Another way to die: autophagic programmed cell death. Cell Death And Differentiation. 2005;12:1528.

(20)

17

22. Evangelisti C, Evangelisti C, Chiarini F, Lonetti A, Buontempo F, Neri LM, et al.

Autophagy in acute leukemias: A double-edged sword with important therapeutic implications.

Biochimica et Biophysica Acta (BBA) - Molecular Cell Research. 2015;1853(1):14-26.

23. Mrakovcic M, Fröhlich LF. p53-Mediated Molecular Control of Autophagy in Tumor Cells.

Biomolecules. 2018;8(2):14.

24. Crighton D, Wilkinson S, O'Prey J, Syed N, Smith P, Harrison PR, et al. DRAM, a p53- Induced Modulator of Autophagy, Is Critical for Apoptosis. Cell. 2006;126(1):121-34.

25. Maiuri MC, Galluzzi L, Morselli E, Kepp O, Malik SA, Kroemer G. Autophagy regulation by p53. Current opinion in cell biology. 2010;22(2):181-5.

26. Xu-Monette ZY, Medeiros LJ, Li Y, Orlowski RZ, Andreeff M, Bueso-Ramos CE, et al.

Dysfunction of the TP53 tumor suppressor gene in lymphoid malignancies. Blood.

2012;119(16):3668.

27. Kamenetsky M, Middelhaufe S, Bank EM, Levin LR, Buck J, Steegborn C. Molecular

details of cAMP generation in mammalian cells: a tale of two systems. Journal of molecular biology.

2006;362(4):623-39.

28. Rosenbaum DM, Rasmussen SG, Kobilka BK. The structure and function of G-protein- coupled receptors. Nature. 2009;459(7245):356-63.

29. Brudvik KW, Tasken K. Modulation of T cell immune functions by the prostaglandin E(2) - cAMP pathway in chronic inflammatory states. British journal of pharmacology. 2012;166(2):411-9.

30. McAllister SS, Weinberg RA. The tumour-induced systemic environment as a critical regulator of cancer progression and metastasis. Nat Cell Biol. 2014;16(8):717-27.

31. Shain KH, Dalton WS, Tao J. The tumor microenvironment shapes hallmarks of mature B- cell malignancies. Oncogene. 2015;34(36):4673-82.

32. Li HJ, Reinhardt F, Herschman HR, Weinberg RA. Cancer-stimulated mesenchymal stem cells create a carcinoma stem cell niche via prostaglandin E2 signaling. Cancer discovery.

2012;2(9):840-55.

33. Aggarwal S, Pittenger MF. Human mesenchymal stem cells modulate allogeneic immune cell responses. Blood. 2005;105(4):1815.

34. Zhi-Gang Z, Wei-Ming L, Zhi-Chao C, Yong Y, Ping Z. Immunosuppressive properties of mesenchymal stem cells derived from bone marrow of patient with hematological malignant diseases. Leukemia & lymphoma. 2008;49(11):2187-95.

35. Kamdje AHN, Krampera M. Notch signaling in acute lymphoblastic leukemia: any role for stromal microenvironment? Blood. 2011;118(25):6506.

36. Battula VL, Chen Y, Cabreira MdG, Ruvolo V, Wang Z, Ma W, et al. Connective tissue growth factor regulates adipocyte differentiation of mesenchymal stromal cells and facilitates leukemia bone marrow engraftment. Blood. 2013;122(3):357.

37. Konopleva MY, Jordan CT. Leukemia stem cells and microenvironment: biology and therapeutic targeting. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2011;29(5):591-9.

38. Naderi E, Skah S, Ugland H, Myklebost O, Lise Sandnes D, Torgersen M, et al. Bone marrow stroma-derived PGE2 protects BCP-ALL cells from DNA damage-induced p53 accumulation and cell death2015. 14 p.

39. Naderi EH, Skah S, Ugland H, Myklebost O, Sandnes DL, Torgersen ML, et al. Bone marrow stroma-derived PGE2 protects BCP-ALL cells from DNA damage-induced p53 accumulation and cell death. Molecular Cancer. 2015;14(1):14.

40. Kastenhuber ER, Lowe SW. Putting p53 in Context. Cell. 2017;170(6):1062-78.

41. Adiseshaiah PP, Skoczen SL, Rodriguez JC, Potter TM, Kota K, Stern ST. Autophagy Monitoring Assay II: Imaging Autophagy Induction in LLC-PK1 Cells Using GFP-LC3 Protein Fusion Construct. Methods in molecular biology (Clifton, NJ). 2018;1682:211-9.

(21)

18

42. Cmhgallo. Schematic representing the formation of an autolysosome and the points of intervention of bafilomycin A1 wikipedia2018 [Available from:

https://commons.wikimedia.org/wiki/File:Autophagy_bafilomycin.png.

43. Insel PA, Ostrom RS. Forskolin as a tool for examining adenylyl cyclase expression, regulation, and G protein signaling. Cellular and molecular neurobiology. 2003;23(3):305-14.

44. Skah S, Richartz N, Duthil E, Gilljam KM, Bindesboll C, Naderi EH, et al. cAMP-mediated autophagy inhibits DNA damage-induced death of leukemia cells independent of p53. Oncotarget.

2018;9(54):30434-49.

45. Gottlieb RA, Andres AM, Sin J, Taylor DPJ. Untangling autophagy measurements: all fluxed up. Circulation research. 2015;116(3):504-14.

46. Skah S, Richartz N, Duthil E, Gilljam KM, Bindesbøll C, Naderi EH, et al. cAMP-mediated autophagy inhibits DNA damage-induced death of leukemia cells independent of p53. Oncotarget.

2018;9(54):30434-49.

47. Orhon I, Reggiori F. Assays to Monitor Autophagy Progression in Cell Cultures. Cells.

2017;6(3):20.

48. Roberts EA, Deretic V. Autophagic proteolysis of long-lived proteins in nonliver cells.

Methods in molecular biology (Clifton, NJ). 2008;445:111-7.

49. Bustin SA, Nolan T. Pitfalls of quantitative real-time reverse-transcription polymerase chain reaction. Journal of biomolecular techniques : JBT. 2004;15(3):155-66.

(22)

Oncotarget 30434

www.oncotarget.com

cAMP-mediated autophagy inhibits DNA damage-induced death of leukemia cells independent of p53

Seham Skah^1,*, Nina Richartz^1,*, Eva Duthil¹, Karin M. Gilljam¹, Christian Bindesbøll¹, Elin Hallan Naderi², Agnete B. Eriksen¹, Ellen Ruud^3,4, Marta M. Dirdal³, Anne Simonsen^1,5 and Heidi Kiil Blomhoff¹

1Department of Molecular Medicine, Institute of Basic Medical Sciences, University of Oslo, Oslo, Norway 2Department of Oncology, Section for Head and Neck Oncology, Oslo University Hospital, Oslo, Norway

3Department of Hematology and Oncology, Division of Pediatric and Adolescent Medicine, Oslo University Hospital, Oslo, Norway

4Institute of Clinical Medicine, Faculty of Medicine, University of Oslo, Oslo, Norway

5Centre for Cancer Cell Reprogramming, Institute of Clinical Medicine, Faculty of Medicine, University of Oslo, Oslo, Norway

*These authors have contributed equally to this work

Correspondence to: Heidi Kiil Blomhoff, email: [email protected] Keywords: cAMP-signaling; autophagy; DNA damage; p53; apoptosis

Received: March 14, 2018 Accepted: June 23, 2018 Published: July 13, 2018

Copyright: Skah et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License 3.0 (CC BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

ABSTRACT

Autophagy is important in regulating the balance between cell death and survival, with the tumor suppressor p53 as one of the key components in this interplay. We have previously utilized an in vitro model of the most common form of childhood cancer, B cell precursor acute lymphoblastic leukemia (BCP-ALL), to show that activation of the cAMP signaling pathway inhibits p53-mediated apoptosis in response to DNA damage in both cell lines and primary leukemic cells. The present study reveals that cAMP-mediated survival of BCP-ALL cells exposed to DNA damaging agents, involves a critical and p53-independent enhancement of autophagy. Although autophagy generally is regarded as a survival mechanism, DNA damage-induced apoptosis has been linked both to enhanced and reduced levels of autophagy. Here we show that exposure of BCP-ALL cells to irradiation or cytotoxic drugs triggers autophagy and cell death in a p53-dependent manner. Stimulation of the cAMP signaling pathway further augments autophagy and inhibits the DNA damage-induced cell death concomitant with reduced nuclear levels of p53. Knocking-down the levels of p53 reduced the irradiation-induced autophagy and cell death, but had no effect on the cAMP-mediated autophagy. Moreover, prevention of autophagy by bafilomycin A1 or by the ULK-inhibitor MRT68921, diminished the protecting effect of cAMP signaling on DNA damage-induced cell death. Having previously proposed a role of the cAMP signaling pathway in development and treatment of BCP-ALLs, we here suggest that inhibitors of autophagy may improve current DNA damage-based therapy of BCP-ALL - independent of p53.

INTRODUCTION

Improved awareness of the vital cellular process of autophagy has in recent years enhanced our understanding

of cancer development as well as mechanisms underlying resistance to cancer treatment [1–4]. Macroautophagy, hereafter referred to as autophagy, involves bulk degradation of cytoplasmic components like damaged

www.oncotarget.com Oncotarget, 2018, Vol. 9, (No. 54), pp: 30434-30449 Research Paper

(23)

Oncotarget 30435

www.oncotarget.com

organelles and long-lived proteins. A double-membraned vesicle, the autophagosome, forms as it sequesters cargo destined for degradation, and the content is degraded after fusion between the autophagosomes and lysosomes [5, 6]. Multiple key proteins have been implicated in the various steps of the autophagic process, including Unc- 51 like autophagy activating kinase (ULK1) involved in the early steps of autophagophore formation [7], and the microtubule-associated protein1 (MAP1) light chain 3 (LC3) widely used as a marker for assessing autophagic flux [8, 9].

Autophagy is required for cells and tissues to maintain homeostasis at critical times of energy demand and cellular stress, and it is considered to be an important regulator of the balance between cell death and cell survival [10–12]. Although autophagy is generally regarded as a survival mechanism, extensive autophagy has also been linked to cell death. Numerous studies have shown that autophagy may either promote or prevent cell death in response to DNA damage [10, 13, 14]. Most studies however, conclude that inhibition of autophagy results in enhanced DNA damage-induced apoptosis, supporting a protective role for autophagy in the DNA damage response (DDR) [13].

In the present study, we reveal the interplay between DDR, p53, apoptosis, and autophagy in leukemia cells.

B-cell precursor acute lymphoblastic leukemia (BCP- ALL) is the most common form of pediatric cancers [15].

Despite the general favorable survival rate of children with BCP-ALL, there is ongoing research to improve the treatment efficiency of subgroups with poor prognosis [15]. The poor prognostic group of BCR/ABL1 positive BCP-ALLs appears particularly dependent on autophagy for their survival and malignant transformation [16].

Treatment of lymphoid malignancies with DNA damaging anti-cancer agents will induce cell cycle arrest, DNA repair, apoptosis or autophagy depending on the balance between these processes. A higher level of autophagy is generally associated with worse clinical outcome [17]. In line with this notion, it has been reported that inhibition of autophagy overcomes treatment resistance in lymphoid malignant cells [18].

There are multiple mechanisms proposed to explain how DNA damage promotes autophagy, including activation of ataxia-telangiectasia mutated (ATM) [19]

and induction of nuclear p53 [20, 21]. BCP-ALLs develop in the bone marrow in close contact with stromal cells that produce prostaglandin E2 (PGE2) [22], and BCP- ALL cells also express functional PGE2 receptors (EP2) [23]. We have previously shown that PGE2 produced by residential stromal cells in the bone marrow limits DNA damage-induced p53 levels via activation of the cAMP signaling pathway, and we have proposed that this may have detrimental effects on both development and treatment of BCP-ALL [24]. Thus, we have shown that cAMP signaling inhibits p53-mediated apoptosis of BCP-

ALL cells exposed to irradiation or cytotoxic drugs [24–

26]. Here, we have uncovered a novel p53-independent link between cAMP-mediated enhancement of autophagy and its ability to reduce DNA damage-induced apoptosis in BCP-ALL cells.

RESULTS

cAMP signaling enhances autophagy induced by DNA damaging agents in REH cells

We have previously shown that activation of the cAMP signaling pathway limits DNA damage-induced apoptosis in BCP-ALL cell lines as well as in primary leukemic cells [24, 26, 27]. Here we aimed to elucidate whether cAMP-mediated survival of BCP-ALL cells involves enhanced autophagy. To test this hypothesis, we treated the ALL cell line REH with the adenylate cyclase activator forskolin, at the optimal concentration of 60 μM, followed by X-ray-mediated irradiation (IR) at 10 Gy. To measure autophagic flux we took advantage of the well-established marker of phagophores and autophagosomes, LC3B. Upon induction of autophagy, the cytosolic form of LC3B (LC3-I) becomes conjugated to phosphatidylehtanolamine (PE) in phagophore membrane and converted to LC3-II. Because the two forms run at different molecular weights when analyzed by western blotting, the LC3-II/I ratio normalized to loading control is therefore commonly used to assess the formation of autophagosomes [8, 9]. As shown in Figure 1A, both IR and forskolin alone induced autophagosome accumulation as assessed by the enhanced LC3-II/LC3-I ratio. The effect of forskolin on LC3-II formation was stronger than that of IR alone and was notable after 6 hours, but more pronounced after 24 hours. Forskolin markedly enhanced the IR-induced LC3-II/I ratio, most prominent after 24 hours.

Accumulation of autophagosomes can be the result of either induced formation of autophagosomes (induced autophagic flux) or be due to blocked autophagosome degradation [8]. To distinguish between these two possibilities, the same experiments were performed in the presence of the lysosomal inhibitor bafilomycin A1 (BafA1). BCP-ALL cells are known to be sensitive to BafA1-treatment [28], and dose response experiments revealed that 2 nM of BafA1 was the optimal non-toxic concentration for REH cells (data not shown). As shown in Figure 1A, the LC3-II/I ratios induced by IR and/or forskolin were clearly enhanced by BafA1 - suggesting enhanced autophagic flux. In Figure 1A, BafA1 was added from the start of the culture. However, to avoid adverse effects of the inhibitor, we also assessed the LC3-II/I ratios after shorter exposure to BafA1. As shown in the left panel of Figure 1B, we concluded that it was sufficient with 2 nM of BafA1 for the last 4 hours prior to cell harvesting. When using these conditions,

(24)

Oncotarget 30436

www.oncotarget.com

we found that forskolin significantly (p<0.01) enhanced the IR-induced LC3-II/I ratio from 4.95 to 9.78 (Figure 1B, right panel). Taken together, we have shown that forskolin and IR independently induces autophagy, and that forskolin is able to potentiate the irradiation-induced autophagy.

Protein kinase a mediates the effects of forskolin cAMP signaling induced by forskolin may result in activation of different effector molecules, including protein kinase A (PKA), Epac and cyclin nucleotide- gated cation channels [29]. We previously concluded that forskolin-mediated inhibition of DNA damage- induced apoptosis in BCP-ALL cells is mediated via PKA [25]. Here we show that the PKA activator 8-CPT- cAMP induced formation of autophagosomes in the same manner as forskolin – both alone and in the presence of IR (Figure 2A). Furthermore, we showed that the PKA inhibitor RP-8-Br-cAMP reduced the forskolin-mediated enhancement of IR-induced autophagy (Supplementary

Figure 1A), and that the phosphodiesterase inhibitor IBMX enhanced the effects of low concentrations of forskolin on autophagy (Supplementary Figure 1B).

Autophagy was here quantified by staining the cells with a newly developed dye CYTO-ID, reported to selectively stain autophagocytic vesicles [30]. We also demonstrated that the potentiating effects of cAMP signaling on DNA damage-induced autophagosome formation in REH cells was not limited to IR, but that forskolin also enhanced the LC3-II/I ratio induced by other DNA damaging agents, such as the leukemia relevant drug doxorubicin (Figure 2B).

cAMP signaling increased the autophagic flux in REH cells

Having demonstrated that cAMP signaling enhances LC3-II formation both alone and in the presence of DNA damaging agents, we next confirmed the formation of autophagosomes by assessing LC3-II puncta by confocal microscopy. As shown in Figure 3, forskolin and IR

Figure 1: cAMP signaling enhances the DNA damage-induced LC3-II/LC3-I ratio. (A and B) REH cells (0.6x10⁶ cells/ml) were incubated in the presence or absence of forskolin (Forsk, 60 μM) for 45 min prior to irradiation (IR, 10Gy), and total lysates were subjected to immunoblot analyses with antibodies against LC3B or calnexin (CANX). The numbers indicated below the LC3 images represent the LC3-II/LC3-I signal ratios relative to the CANX signals, normalized to the ratio in untreated (Ctrl) cells. (A) When indicated, BafA1 (2 nM) was added from the start of the cultures. The cells were harvested at the indicated time points after IR, and one representative Western blot of three is shown. (B) Left panel: The cells were harvested 24 hours after IR, and BafA1 (2 nM) was added for the last 4 hours, as indicated. One representative Western blot of 8 is shown. Right panel: Ratios of the LC3-II/LC3-I signal intensities relative to the CANX signals, normalized to ratio in untreated (Ctrl) cells. The data represent the mean +/- SEM, n=8. ^*p< 0.05 (paired t test).

(25)

Oncotarget 30437

www.oncotarget.com

independently increased the number and sizes of LC3- II puncta after 24 hours, with enhanced levels when the two treatments were combined. We further confirmed the induction of autophagy by staining the cells with CYTO- ID. In Figure 4A, we show CYTO-ID staining of REH cells treated with IR in the presence of forskolin, as revealed by confocal microscopy. The co-localization between CYTO-ID staining and LC3-II puncta is demonstrated in Figure 4B. We demonstrated that pre-incubating the cells for 30 min with the ULK1 inhibitor MRT68921 at the optimal concentration of 100 nM prevented the forskolin- induced CYTO-ID staining as assessed by flow cytometry (Figure 4C). The same effect was observed with siRNA against ULK1 (see Supplementary Figure 2). Careful kinetic experiments concluded that optimal CYTO-ID staining was obtained between 12 and 24 hours, with a clear induction by forskolin noted already after 6 hours of

treatment (Figure 4D). Treatment with BafA1 augmented the CYTO-ID staining measured after 24 hours (Figure 4E), enhancing the fold induction of IR-induced CYTO-ID fluorescence intensity from approximately 2.5 to 4. Thus, again we concluded that cAMP signaling enhances DNA damage-induced autophagy.

To further support the cAMP-mediated enhancement of IR-induced autophagy, we measured autophagic flux as the degradation of long-lived proteins, known to be mainly degraded by autophagy [9]. Accordingly, IR alone enhanced the degradation of long-lived proteins in REH cells, and forskolin significantly (p=0.01) enhanced this degradation (Figure 4F).

Autophagy-related genes (ATGs) are differentially regulated at transcriptional, post-transcriptional and post- translational levels [31]. Since IR has been shown to induce transcription of the LC3B-coding gene MAP1LC3B

Figure 2: PKA- and doxorubicin-mediated autophagy. (A and B) REH cells were treated with or without forskolin, IR and BafA1 as described in Figure 1B. When indicated, the cells were treated with or without 8CPT-cAMP (8CPT, 200μM) 45 min prior to IR (panel A) or with 150 nM doxorubicin (Doxo) 45 min after adding forskolin (panel B). Left panels: One representative Western blot of three independent experiments is shown. The numbers indicated below the LC3 images represent the LC3-II/LC3-I signal ratios relative to the CANX signals, normalized to the ratio in untreated (Ctrl) cells. Right panels: Ratios of the LC3-II/LC3-I signal intensities relative to the CANX signals, normalized to the ratio in untreated (Ctrl) cells. The data represent the mean +/- SEM, n=3. ^*p<0.05 (paired t test).

(26)

Oncotarget 30438

www.oncotarget.com

[32], we performed qRT-PCR of this gene in REH cells. As shown in Supplementary Figure 3, IR and forskolin alone produced only marginally elevated levels of MAP1LC3B mRNA. However, clear additive effects on MAP1LC3B mRNA levels were obtained when combining the two treatments. The total level of LC3B protein (LC3-I + LC3- II) was not enhanced in REH cells co-treated by IR and forskolin as compared to control (see Figure 1).

Forskolin enhances DNA damage-induced autophagy in NALM-6 and primary BCP-ALL cells

We have previously shown that cAMP signaling regulates DNA damage-induced apoptosis in a similar manner in REH cells, NALM-6 and in primary leukemic cells from patients with BCP-ALL [24]. We therefore investigated whether cAMP signaling also enhanced the IR-induced autophagy in NALM-6, and in primary leukemic cells, using cells from three patients with

BCP-ALL. Indeed, the CYTO-ID fluorescence intensity increased in NALM-6 cells when treated with IR in the presence or absence of forskolin (Figure 5A), as was also the case for cells derived from three patients with BCP-ALL (Figure 5B). Due to limited number of cells, we did not assess the ability of forskolin alone to induce autophagy in cells obtained from patient #1 and #2.

Autophagy is involved in cAMP-mediated survival of DNA damaged cells

Having established the ability of forskolin to enhance the level of DNA damage-induced autophagy, we aimed to identify a possible link between the increased autophagy and the reduced cell death promoted by cAMP signaling in REH cells. As shown in Figure 6A, IR alone induced cell death in approximately 21% of the cells as measured after 24 hours, and in 59% after 48 hours. In line with our previous results [24], forskolin alone had only minor effects on the basal levels of cell death in

Figure 3: Immunocytochemistry of LC3-puncta. (A and B) REH cells were treated with or without forskolin, IR and BafA1 as described in Figure 1B, with addition of BafA1 to all samples. The cells were subjected to immunocytochemistry for the detection of LC3 puncta by confocal microscopy, and the cells were co-stained with Hoechst for visualization of the nuclei. (A) One representative of three independent experiments is shown. Scale bars = 10μm. (B) The number of LC3 puncta per cell from three independent experiments were quantified, counting at least 30 cells. The data represent the mean +/- SEM, n = 30. The numbers of small and large puncta are indicated.

(27)

Oncotarget 30439

www.oncotarget.com

REH cells, but significantly reduced the IR-induced cell death after 24 hours and 48 hours. To investigate the link between forskolin-induced autophagy and increased survival of the DNA damaged cells, we used inhibitors of autophagy at doses that were not toxic to the cells after 48 hours of treatment, but still retained the ability to prevent autophagic degradation. By using BafA1 at 2 nM from start of the cultures, we found that the forskolin- mediated protection of cell death was significantly reduced after both 24 hours and 48 hours (Figure 6A).

The same tendency was observed when treating the cells with 5μM of chloroquine (data not shown). Finally, we demonstrated that inhibiting autophagy by the ULK1 inhibitor MRT68921 impeded the cAMP-mediated protection against DNA damage-induced cell death after both 24 hours and 48 hours (Figure 6B).

The involvement of p53 in IR-induced autophagy The tumor suppressor p53 has been implicated in regulation of autophagy, in particular related to cellular stress [17, 20]. It is generally believed that nuclear levels of p53 promote autophagy, whereas cytosolic levels prevent the autophagy process [20, 33]. Having

previously established that cAMP-mediated inhibition of DNA damage-induced apoptosis of BCP-ALL cells involves down-regulation of p53 [24, 25], we here confirmed the ability of forskolin to reduce the level of IR-induced p53 commencing as early as 4 hours after IR (Supplementary Figure 4A). In order to assess whether the subcellular localization of p53 was affected by any of the treatments, we performed confocal microscopy of REH cells stained with an antibody directed against p53. As shown in Figure 7A, we found that IR enhanced both the nuclear and cytosolic levels of p53, whereas co-treatment with forskolin selectively reduced the levels of p53 in the nuclei. Forskolin alone had no effect on the nuclear localization of p53 (Figure 7A and 7B). To confirm these findings, we performed cellular fractionation experiments followed by immunostaining of p53. The data presented in Figure 7C confirm that forskolin selectively inhibits accumulation of p53 within the nuclei of irradiated REH cells.

To unravel the link between p53 and autophagy in our experimental settings, p53 was targeted by siRNA.

The knock-down of p53 by siRNA is demonstrated by Western blot analysis in Figure 8A. Supporting a stimulatory role in autophagy, siRNA against p53 reduced

Figure 4: cAMP signaling enhances autophagic flux in REH cells. (A-E) REH cells were treated with or without forskolin, IR and BafA1 as described in Figure 1B. CYTO-ID staining was performed 24 hours after IR – if not otherwise indicated, and the staining intensity was analyzed by flow cytometry and normalized to untreated (Ctrl) cells. (A) Confocal images of IR/forskolin-treated cells stained with CYTO-ID, (B) The co-localization between the autophagosomal marker CYTO-ID and LC3-puncta was analyzed by confocal imaging of IR/forskolin-treated cells. (C) The cells were pretreated with the ULK1 inhibitor MRT68921 (100 nM) for 30 min prior to adding forskolin, and the cells were irradiated after another 45 min. The data represent the mean CYTO-ID fluorescence intensity +/- SEM, n=3.

*p=0.05 (paired t test). (D) The cells were stained with CYTO-ID at the indicated time points, and the fluorescence intensity was analyzed by flow cytometry. The data represent the mean CYTO-ID fluorescence intensity +/- SEM, n=5. ^*p<0.05 (paired t test). (E) Cells were treated with or without 2 nM BafA1for the last 4 hours of the 24 hours incubation. The data represent the mean CYTO-ID fluorescence intensity +/- SEM, n=5. ^*p<0.05 (paired t test). (F) The effect of IR and forskolin on relative autophagic flux was quantified by measuring the degradation of long-lived proteins as described in Materials and Methods. The data represent the mean +/- SEM, n=3, and the values are normalized to the degradation in untreated (Ctrl) cells. ^*p<0.05 (paired t test).

(28)

Oncotarget 30440

www.oncotarget.com

the IR-induced autophagy as revealed by the reversion of the LC3 II/I ratio (Figure 8B and 8C) and by CYTO-ID staining (Figure 8D). Furthermore, siRNA also reduced the IR-mediated cell death (Supplementary Figure 4, panel B). However, the cAMP-mediated enhancement of IR-induced autophagy could not be explained by changed localization of p53. First of all, the nuclear and not the cytosolic levels of p53 were reduced by the co-treatment with forskolin (see Figure 7). Secondly, siRNA against p53 had no effect on the ability of forskolin to enhance the IR-induced autophagy (Figure 8B-8D).

DISCUSSION

We have previously established an in vitro model of BCP-ALL, successfully used for studying the interplay between p53 levels and DNA damage-induced cell death related to development and treatment of this disease [24–

27]. In the present study, we extend this model to unravel a novel p53-independent interplay between autophagy and cell death with implications for treatment of BCP-ALL.

Suppression of normal p53 functions is regarded as a prerequisite for the development of most cancers [34]. Thus, mutations in the TP53 gene itself or in p53- regulating genes render the malignant cells resistant to control mechanisms that are part of the normal DNA damage response [35]. As most childhood BCP- ALLs retain wild type TP53 at diagnosis [36], one may assume that these leukemic cells depend on alternative strategies to mitigate the function of wild type p53. We have previously suggested that stimulation of the cAMP signaling pathway may represent such a mechanism, since elevated levels of cAMP in BCP-ALL blasts suppress DNA damage-induced p53 accumulation and apoptosis [26]. Here we demonstrate that cAMP signaling also enhances DNA damage-induced autophagy in BCP- ALL blasts, enabling us to reveal the interplay between autophagy and apoptosis in these cells, and to dissect the role of p53 in these processes.

In most cell types, DNA damage will induce both autophagy and apoptosis [14]. However, there is no consensus as to whether the induced autophagy is required

Figure 5: cAMP signaling enhances CYTO-ID staining in NALM-6 and in primary BCP-ALL cells. (A) NALM-6 cells (0.6x10⁶ cells/ml) were treated with or without forskolin (Forsk, 60μM) for 45 min prior to irradiation (IR, 5Gy). BafA1 (2 nM) was added to the cell cultures for the last 4 hours of the 24 hours incubation, before the cells were stained with CYTO-ID and analyzed for fluorescence intensity by flow cytometry. The data represent the mean CYTO-ID fluorescence intensity ± SEM, n=4. ^*p<0.05 (paired t test). (B) Primary leukemic blasts (0.6 x10⁶ cells/ml) from three patients diagnosed with BCP-ALL were treated with or without Forsk, IR and BafA1 as described in panel A.

(29)

Oncotarget 30441

www.oncotarget.com

for the apoptosis or actually has a protective role [10, 13, 14]. We found that treatment of BCP-ALL cells with IR or doxorubicin promoted autophagy and killed the cells.

These effects were notable in the cell line REH, as well as in primary leukemic blasts isolated from children with BCP-ALL. Blocking autophagy by BafA1 or the ULK- inhibitor MRT68921 had little or no effect on the IR-

induced killing of the BCP-ALL cells. Thus, although increased autophagy has been linked to unfavorable clinical outcome of DNA damaging cancer treatments of patients with lymphoid malignancies [17], our results suggest that the level of autophagy induced by DNA damaging agents in vitro is too low to protect the BCP- ALL cells from the lethal DNA lesions. It was therefore

Figure 6: cAMP-mediated inhibition of DNA damage-induced cell death involves autophagy. REH cells were treated with or without forskolin and irradiation as described in Figure 1. When indicated, BafA1 (2 nM) (A) or the ULK1 inhibitor MRT68921 (100 nM) (B) was present in the cell cultures throughout the experiments to block autophagy. The percentage of PI-positive cells was analyzed by flow cytometry 24 hours or 48 hours after IR, as indicated. The results are presented as the mean ± SEM, n=6. ^*p<0.05 (paired t test).

cAMP-mediated autophagy inhibits DNA damage-induced death of leukemia cells independent of p53-article summary

cAMP-mediated autophagy inhibits DNA damage-induced death of

leukemia cells independent of p53

Article summary

Eva Duthil

Project thesis ved Institutt for medisinske basalfag UNIVERSITETET I OSLO

ABSTRACT

Table of Contents

1. INTRODUCTION

2. MOTIVATION AND AIMS OF THE PRESENT PROJECT

3. RESULTS

4. DISCUSSION

5. CONCLUSIONS

6. REFERENCES

cAMP-mediated autophagy inhibits DNA damage-induced death of leukemia cells independent of p53